| Περίληψη: | Manufacturing Systems and Automation has faced a rapid evolution/digital transformation during the fourth industrial revolution or Industry 4.0. This thesis embarks on an investigation of the trends and advancements in the field of manufacturing systems and the underlying algorithms over the last decade, with the focal point being the Machine Learning (ML) and Artificial Intelligence (AI) frameworks. Furthermore, the shift of manufacturing systems towards automation gave birth to a humongous volume of data. Thus, techniques to utilize and organize all this data for the benefit of contemporary industries, using smart manufacturing systems, are proposed in this thesis. Furthermore, a core element of this diploma thesis is Machine Vision (MV). Initially, the definition of MV, is provided the fundamental methods developers should follow, in order to build a proper MV system are analyzed. Additionally, examples of MV frameworks in modern industry are displayed and analyzed. Moreover, with ML and AI being in the core of this research, following their definitions, applications in manufacturing. frameworks, architectures and systems based on ML in industrial case studies are then presented. Moreover, this research looks into Flexible Manufacturing Cells (FMCs) and their employment in modern manufacturing. Following the definition of FMCs, their Advantages and Disadvantages are presented together with the FMCs’ component analysis. Lastly, as in the modern industry application of FMCs the use of Computer-Aided-Design (CAD) is of critical importance, hence current advantages of CAD files and ways to utilize their data, to contribute to the automation of the FMCs effectively and efficiently, are researched and proposed within this diploma thesis. In conclusion, this thesis essentially probes into the Process planning of FMCs and a framework that promotes strategies and technologies to achieve the desired data handling techniques and taxonomy in FMCs, is proposed. Ultimately, the results of the proposed framework, as well as the conclusions, and future work, based on this thesis implementations, are handed out and further analyzed.
|
|---|
